The main aim of agencies involved in the construction of asphalt roads is to improve the field performance of the asphalt mixtures. The rising use of recycled and novel materials in asphalt mixture has rendered the previous semi-empirical methods of mixture design partly incapable of accurately predicting the mixture field performance with high precision. Meeting this challenge calls for a shift towards an approach involving mixture performance tests. This project deals with investigating the performance of modern recycled asphalt mixes containing ground tire rubber, Recycled Asphalt Shingles (RAS), Recycled Asphalt Pavement (RAP) and rejuvenators. Various performance tests for various type of distresses were considered to evaluate the effect of using these components in asphalt mixtures. Combining these performance tests with prediction of field performance of mixtures should provide more robust and reliable design criteria for the modern recycled asphalt mixtures leading to better roads. To this end, the performance of eighteen different dense-graded asphalt mixtures paved in Missouri were investigated. The sections contain a wide range of reclaimed asphalt pavement (RAP) and recycled asphalt shingles (RAS), and different types of additives. The large number of sections investigated and the associated breadth of asphalt mixtures tested provided a robust data set to evaluate the range, repeatability, and relative values provided by modern mixture performance tests. As cracking is one of the most prevalent distresses in Missouri, performance tests such as the disk-shaped compact tension test (DC[T]) and Illinois flexibility index test (I-FIT) were used to evaluate the cracking potential of the sampled field cores. In addition, the Hamburg wheel tracking test (HWTT) was employed to assess rutting and stripping potential. Asphalt binder replacement (ABR) and binder grade bumping at low temperature were found to be critical factors in low-temperature cracking resistance as assessed by the DC(T) fracture energy test. Six sections were found to perform well in the DC(T) test, likely as a result of binder grade bumping (softer grade selection) or because of low recycling content. However, all of the sections were characterized as having brittle behavior as predicted by the I-FIT flexibility index. Service life and ABR were key factors in the I-FIT test. Finally, a performance-space diagram including DC(T) fracture energy and HWTT depth was used to identify mixtures with higher usable temperature interval (UTI mix), some of which contained significant amounts of recycled material. In the second phase of chapter 2, the poor performing mixtures were redesigned in order to improve their performance by changing the components of the mixtures including recycling content, rejuvenator type and amount, binder type, crumb rubber quantity, etc.. Finally, the optimum content of the components based on mixture performance and materials costs was determined. The testing results along with the field performance data was used to develop a specification for MoDOT to screen the mixtures and use it for quality control and quality assurance of plant-produced asphalt concrete. Field monitoring is a potential means to identify the most reliable cracking performance test. Also, a new cracking index was introduced based on SCB (I-FIT) test to improve the test reliability and correlation with field results. In the third chapter of this study a prediction tool was developed to predict the performance of asphalt mixture at high and low temperatures. This tool is based on two different prediction models for DC(T) fracture energy and Hamburg wheel track tests. For DC(T) fracture energy model, genetic programming was used to develop the prediction model, and Convolution Neural Network (CNN) was used to train the Hamburg wheel track model on 10,000 data points. A database containing a comprehensive collection of Hamburg and DC(T) tests results were used to develop the machine learning-based prediction models. This tool can be used for pre-design purposes to design an asphalt mixture with balanced performance in rutting and cracking. The models were formulated in terms of typical influencing mixture properties variables such as asphalt binder high-temperature performance grade (PG), mixture type, aggregate size, aggregate gradation, asphalt content, total asphalt binder recycling content and tests parameters like temperature and number of cycles. Models accuracy were assessed through a rigorous validation process and found to be quite acceptable, despite the relatively small size of the training set. Since performing performance tests might be cost-restrictive for some users, using the proposed ML-based models can save time and expense during the material screening phase. Pavement distress inspections are performed using sophisticated data collection vehicles and/or foot-on-ground surveys. In either approach, the process of distress detection is human-dependent, expensive, inefficient, and/or unsafe. Automated pavement distress detection via road images is still a challenging issue among pavement researchers and computer-vision community. In the forth chapter of dissertation, we extracted 7237 google street-view, manually annotated for classification (nine categories of distress classes). Afterward, the YOLO (you look only once) deep learning framework was implemented to train the model using the labeled dataset. Also, U-net based model is developed to quantify the severity of the distresses, and finally, a hybrid model is developed by integrating the YOLO and U-net model to classify the distresses and quantify their severity simultaneously. The output of the distress classification and segmentation models are used to develop a comprehensive pavement condition tool which rates each pavement image according to the type and severity of distress extracted. As a result, we are able to avoid over-dependence on human judgement throughout the pavement condition evaluation process. The outcome of this study could be conveniently employed to evaluate the pavement conditions during its service life and help to make valid decisions for rehabilitation of the roads at the right time.
